TUESDAY BRIEFING: A Safety Spending Test, a $1 Billion Wildfire Fund, and the End of Rule 21's Flat Fee?

Today's briefing examines three CPUC proceedings that could reshape how utilities justify spending and recover costs.

• Commissioner Christine Harada issued a scoping memo for the latest incarnation of the CPUC’s Risk-Based Decision-Making Framework.
• PG&E is seeking to extend a wildfire self-insurance program it says has saved customers $1.8 billion while accumulating more than $1 billion for claims and program expenses.
• And the CPUC is assembling the cost and workload data needed to replace Rule 21’s flat $800 interconnection application fee with a potentially tiered structure.

RISK-BASED DECISION-MAKING FRAMEWORK

Commissioner Christine Harada has scoped the CPUC's new risk-based decision-making rulemaking into three tracks.

Track 1 asks whether the Rate Case Plan should give the CPUC's Safety Policy Division (SPD) more time to evaluate RAMP applications. Track 1 also asks whether Alpine Natural Gas and West Coast Gas should file RSARs on the D.22-10-002 model to comply with Public Utilities Code Section 591. Comments are due October 23 and November 6.

Track 2 reopens the benefit-cost ratio (BCR) methodology adopted in R.20-07-013. The CPUC will consider:

• Whether all utilities should use a single method;
• Whether O&M avoided by an investment belongs in the numerator as a benefit or in the denominator as a cost reduction;
• Whether the cost horizon should run four years or the life of the asset; and
• Whether a BCR of at least 1.0 should be required before a utility makes an investment.

PG&E must file a summary report and workpaper on its Present Value Revenue Requirement (PVRR) multiplier by February 1, 2027, the same day a staff proposal on O&M treatment is scheduled.

Track 3 asks:

• Whether the CPUC should adopt a risk-tolerance standard or a more flexible framework;
• Whether to confine it to catastrophic risks; and
• Whether affordability should figure in the outcome.

The utilities must file a joint survey by July 31, and SPD's proposal is due November 16.

INSTANT ANALYSIS: The central question in Track 2 is whether the CPUC should require a BCR of at least 1.0 before a utility undertakes a risk-mitigation investment. A numerical threshold of that kind would convert what is now a judgment about safety spending into a calculation, and it would make every input to that calculation contestable.

The time-horizon question is the one most likely to change outcomes. Measuring costs over a four-year rate case period produces a different ratio than measuring them over the life of an asset, and capital projects with long payback periods are the ones that would look worse under the shorter measure. PG&E has already adopted a position by using a PVRR multiplier in its RAMP and GRC filings, and Harada has ordered it to explain the mechanics and produce a workpaper.

The CPUC is considering this subject in two places at once. The OIR was served on A.26-02-005, where PG&E, SCE, and SDG&E seek approval of a BCR calculation methodology, an audit methodology and cost recovery conditions under Resolution SPD-37.

On risk tolerance, the memo commits to nothing. It asks whether a standard or a framework is appropriate, whether the inquiry should be limited to catastrophic risks, and whether affordability belongs in the analysis at all.

WILDFIRES

PG&E has asked the CPUC to continue its wildfire liability self-insurance program until the Commission issues a final decision in the utility's pending 2027 General Rate Case.

• The program is authorized through December 31. A proposed decision in the GRC (A.25-05-009) is scheduled for March 2027, with a final decision anticipated in May 2027. Advice Letter 7945-E would bridge that period without modifying the program's existing terms. TURN and Cal Advocates, who negotiated the original self-insurance settlement with PG&E, support the extension.
• PG&E requests authorization by September 30 to give it certainty before the 2023 GRC cycle closes, but separately asks for an effective date on (or before) August 13. Protests are due August 3.
• The settling parties met February 27 to assess the program's first three years, as required by D.23-01-005. All parties agreed the program should continue without modification. The filing discloses that the meeting occurred and that the parties concurred, but not what they reviewed.
• PG&E estimates replacing commercial wildfire liability insurance with self-insurance has saved customers approximately $1.8 billion. PG&E says the self-insurance fund had a balance of about $1.036 billion as of December 31, 2025, available to pay future wildfire liability claims and the program’s operating expenses.

In its 2023 GRC, PG&E sought $707 million annually to purchase $900 million of commercial wildfire liability coverage. PG&E began transitioning to self-insurance in March 2023 and has operated under a fully self-insured structure since August 2023. It reduced the program's revenue requirement for CPUC-jurisdictional customers to zero beginning January 1, 2025, and requests no revenue requirement for the program in the 2027 GRC.

INSTANT ANALYSIS: The extension is procedural and unlikely to draw opposition. The consequential disclosure is the $1.8 billion savings estimate paired with more than $1 billion available for claims, against a 2023 proposal to pay $707 million a year for $900 million of coverage.

The savings estimate is PG&E's own. Nothing in the filing demonstrates the fund would hold up against a catastrophic wildfire. Although the parties met February 27 to evaluate the program’s first three years, the letter reports their conclusion but none of the analysis supporting it.

INTERCONNECTION

A new ALJ ruling orders PG&E, SCE, and SDG&E to produce the cost and workload records the CPUC would need to replace Rule 21's flat $800 interconnection application fee. The Commission believes the fee overcharges simple projects, undercharges complex ones, and may not match the utilities’ aggregate processing costs.

For applications outside the NEM family (which the ruling labels non-NBT/NEM), each utility must:

• Estimate the administrative and engineering hours a review consumes;
• Convert those hours into fully loaded costs; and
• Report the amount over- or under-collected per application and across all applications received last year.

The utilities must also account for how they track staff time, which steps they have automated, how many applications clear initial technical screens, and what conditions send a project into a detailed study. A separate prompt tests the $150 hourly engineering rate carried in the Rule 21 tariff against what utility engineering work actually costs.

The remaining prompts explore whether a fee could vary by capacity, technology, export status and customer class. The ruling asks the utilities to compare non-exporting reviews against limited-exporting reviews, standalone solar against co-located solar and storage, and residential projects against commercial ones.

Appendix A gathers all of it into a single template covering 2024 and 2025 combined, sorted into five capacity bands running from under 30 kW to 5 MW and larger. Each cell calls for applications received, withdrawn and still pending, along with staff hours, actual review cost and aggregate capacity. The utilities must then identify what drives costs upward, with Power Control Systems named as one candidate.

INSTANT ANALYSIS: The Commission has stopped debating whether $800 is equitable and started building the record needed to replace it. A tiered fee would likely lower the cost of entry for small residential projects and raise it for large or export-capable ones.

Whether the record can support that redesign is another question. The ruling accepts estimates wherever cost accounting or time records do not exist, which leaves any proposed tier open to the objection that it rests on utility judgment rather than measurement. The ruling also asks about last year’s applications in one place while combining 2024 and 2025 in the template, so the responses may produce two datasets that do not align.

The template reveals the distinctions around which the CPUC may structure a replacement fee: capacity, technology, export status, and customer class.